Playing some more with this Tweak OpenGL hack, and mixing in Tobias
Germer's GLSL support for Croquet, I now have some simple shaders running. Hooray! As a simple next step, I actually tried to do something that I didn't just copy and assimilate, and of course it doesn't work. I'm wondering if there is something fundamentally wrong with what I'm trying. Borrowing part of the NeHe tutorial number 6, I have added texture coordinates to the simple cube geometry so that I can access those coordinates in the vertex and fragment shaders as the basis for a procedural texture like the simple example towards the bottom of the page at http://croquetconsortium.org/index.php/Procedural_Texturing The main thing that I'm wondering about is: can I use the texture coordinates without there being any actual texture anywhere? Since the color produced in the fragment shader is entirely calculated within the shader, there is no need for an actual texture as far as I am concerned, but will OpenGL let me do this? Also, I would like to make Tobias Germer's GLSL support available again, along with the FBO support that I put together, but these OpenGL extensions are structured as changeSets and I don't think they would work well as Monticello packages, so how should I make them available? I think I asked Josh about this once before; I'll have to go dig out his answer ... When I make them available, these OpenGL extensions may conflict with other similar development that is being done, so use them at your own risk ;-) Any pointers would be appreciated. |
On Jun 11, 2007, at 14:49 , David Faught wrote:
> Playing some more with this Tweak OpenGL hack, and mixing in Tobias > Germer's GLSL support for Croquet, I now have some simple shaders > running. Hooray! > > As a simple next step, I actually tried to do something that I didn't > just copy and assimilate, and of course it doesn't work. I'm > wondering if there is something fundamentally wrong with what I'm > trying. > > Borrowing part of the NeHe tutorial number 6, I have added texture > coordinates to the simple cube geometry so that I can access those > coordinates in the vertex and fragment shaders as the basis for a > procedural texture like the simple example towards the bottom of the > page at > http://croquetconsortium.org/index.php/Procedural_Texturing > > The main thing that I'm wondering about is: can I use the texture > coordinates without there being any actual texture anywhere? Since > the color produced in the fragment shader is entirely calculated > within the shader, there is no need for an actual texture as far as I > am concerned, but will OpenGL let me do this? Yes. > Also, I would like to make Tobias Germer's GLSL support available > again, along with the FBO support that I put together, but these > OpenGL extensions are structured as changeSets and I don't think they > would work well as Monticello packages, so how should I make them > available? I think I asked Josh about this once before; I'll have to > go dig out his answer ... > > When I make them available, these OpenGL extensions may conflict with > other similar development that is being done, so use them at your own > risk ;-) > > Any pointers would be appreciated. Just publish an updated Monticello package. - Bert - |
On 6/11/07, Bert Freudenberg <[hidden email]> wrote:
> On Jun 11, 2007, at 14:49 , David Faught wrote: ... > > Also, I would like to make Tobias Germer's GLSL support available > > again, along with the FBO support that I put together, but these > > OpenGL extensions are structured as changeSets and I don't think they > > would work well as Monticello packages, so how should I make them > > available? I think I asked Josh about this once before; I'll have to > > go dig out his answer ... > > > > When I make them available, these OpenGL extensions may conflict with > > other similar development that is being done, so use them at your own > > risk ;-) > > > > Any pointers would be appreciated. > > Just publish an updated Monticello package. Thanks for your quick reply! Not having the code in front of me here at work, I don't have the proper name, but I'm assuming you mean the package that has the OpenGL class and methods in it. |
On Jun 11, 2007, at 15:21 , David Faught wrote: > On 6/11/07, Bert Freudenberg <[hidden email]> wrote: >> On Jun 11, 2007, at 14:49 , David Faught wrote: > ... >> > Also, I would like to make Tobias Germer's GLSL support available >> > again, along with the FBO support that I put together, but these >> > OpenGL extensions are structured as changeSets and I don't think >> they >> > would work well as Monticello packages, so how should I make them >> > available? I think I asked Josh about this once before; I'll >> have to >> > go dig out his answer ... >> > >> > When I make them available, these OpenGL extensions may conflict >> with >> > other similar development that is being done, so use them at >> your own >> > risk ;-) >> > >> > Any pointers would be appreciated. >> >> Just publish an updated Monticello package. > > Thanks for your quick reply! > > Not having the code in front of me here at work, I don't have the > proper name, but I'm assuming you mean the package that has the OpenGL > class and methods in it. Yes. - Bert - |
On 6/11/07, Bert Freudenberg <[hidden email]> wrote:
> On Jun 11, 2007, at 15:21 , David Faught wrote: > > On 6/11/07, Bert Freudenberg <[hidden email]> wrote: > >> On Jun 11, 2007, at 14:49 , David Faught wrote: > > ... > >> > Also, I would like to make Tobias Germer's GLSL support available > >> > again, along with the FBO support that I put together, but these > >> > OpenGL extensions are structured as changeSets and I don't think > >> they > >> > would work well as Monticello packages, so how should I make them > >> > available? I think I asked Josh about this once before; I'll > >> have to > >> > go dig out his answer ... > >> > > >> > When I make them available, these OpenGL extensions may conflict > >> with > >> > other similar development that is being done, so use them at > >> your own > >> > risk ;-) > >> > > >> > Any pointers would be appreciated. > >> > >> Just publish an updated Monticello package. > > > > Thanks for your quick reply! > > > > Not having the code in front of me here at work, I don't have the > > proper name, but I'm assuming you mean the package that has the OpenGL > > class and methods in it. > > Yes. Hhmm. I notice in the SDK that the OpenGL class is back in the category OpenGL-OpenGL and that all the good stuff is therefore in the Monticello package OpenGL, including the pool dictionaries. This was split out into OpenGL-Pools and OpenGL-Core at one time. So to fix this either a Monticello package named OpenGL-OpenGL should be added and saved in the proper repository, or the category OpenGL-OpenGL should be renamed to OpenGL-Core, right? Which is the right thing? Dave |
Free forum by Nabble | Edit this page |